Identification and Modeling of Vibration Signals in Startup Stage of Electric Driven Reciprocating Pump

Abstract:

How to realize signal modeling and vibration signal characteristic extraction is a very significant topic. A large amount of drilling pump vibration signals were acquired from the indoor tests. The startup signals with time consistency were segmented from these measurement signals and analyzed in detail. There are mainly such five types of vibrations in the startup signals as the pump body’s vibration, whistle, shocks in moving parts, and impacts of the value lifted off or dropped on the seat, friction or grinding between moving parts. The pump body’s vibration and whistle have good time-frequency characteristics and change very regularity, which are defined as the startup vibration in this paper. The pump body’s vibration signals are modeled by OFMM method. After to exclude the OFMM modeling signal, the remaining signal was separated into different integrated components according to their vibration sources by PFM method, a HMM whistle vibration model based on PFM parameters was achieved. Furthermore, A combination of OFMM and HMM model is used to describe the pump startup vibration. Realistic simulation on the pump’s startup vibration has been achieved. Signal simulation was also carried out by use of this combination model. This approach is expected to become a powerful tool for drilling pump’s startup vibration signal analysis and modeling.

• Periodical: Applied Mechanics and Materials (Volumes 184-185)
• Pages: 256-262
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.184-185.256
• Online since: June 2012
• Authors: Miao Rong Lv, Mei Li, Shi Gang Shen, Bao Jian Wei
• Keywords: Pump, Signal Identification, Signal Modeling, Signal Separation, Vibration

References

[1] Guo-feng Wang; Zhi-gao Luo; Xu-da Qin; Yong-gang Leng; Le Chang.Modeling of Nonstationary Signals Based on Time-Varying Autoregression and Its Application in Fault Diagnosis of Bearing[J]. Journal of Tianjin University,2008,41(5):558-562. In Chinese.

[2] Y.Grenier. Time-dependent ARMAM modeling of Nonstationary Signals[J]. IEEE Transactions on Acoustics，Speech，and Signal Processing. August 1983，Vol. ASSP- 31(4):899一911.

DOI: 10.1109/tassp.1983.1164152

[3] Pradip Sircar, Mohanjeet Singh Syali. Complex AM signal model for non-stationary signals. Signal Processing 53 (1996) 35-45.

DOI: 10.1016/0165-1684(96)00074-6

[4] Miaorong Lv, Desheng Gu. The basic principle and method of acoustic signal separation[J]. Journal of Central South University(Science and Technology),2004,35(1): 62—66. In Chinese.

[5] Pradip Sircar, Rakesh K. Saini. Parametric modeling of speech by complex AM and FM signals. Digital Signal Processing 17 (2007) 1055–1064.

DOI: 10.1016/j.dsp.2006.08.011

[6] Miaorong Lv, Desheng Gu. The Primary Study for Application of Optimum Matching Frequency Method in Signal Processing[J]. Journal of Central South University(Science and Technology), 2003,34(2):196~199. In Chinese.

[7] Miaorong Lv, Qian Wang.Application of Pattern Filter Method on Well Site Vibration Signal Separation[J]. Noise and Vibration Control,2010,30(2):107-110. In Chinese.

[8] Miaorong Lv, Lin Zhou, Li Wang. A New Method of Quasi-Periodic Signal Processing in Petroleum Engineering[J]. Petroleum Drilling Techniques, 37(5):89-92. In Chinese.